The health sector requires

computerized work in technology, as the world is

developing rapidly in science and technology and

made a lot of progress in everything. The most

challenging tasks in the medical field are heart

stroke, recently. By understanding the risk

factors for heart attack will reduces the severity

and increases patient's health. At present, for

every one minute one person is dies because of

heart Stroke. The stroke treatment needs to start

as soon as possible, when the patient's prognosis is

known. In earlier days, to predict strokes machine

learning (ML) algorithms are used. The risk of

heart disease and stroke is predicted in this study

uses an attribute-based machine learning

technique. So, in this analysis Cleveland heart

dataset's full set of attributes are used. The three

attribute evaluators are used in this paper to train

XGboost classifiers for effective heart

disease/stroke risk predictions are collected from

optimal attribute sets. The performance has been

compared using parameters like Accuracy,

precision, recall and F1-score. The results verify

that XGboost algorithm has achieved the highest

performance parameter values (Accuracy as 98%,

Precision as 97%, Recall as 97.5% and F1-Score

as 98%) compared to other ML algorithms.